[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-93266":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":14,"stars7d":15,"stars30d":15,"stars90d":14,"forks30d":14,"starsTrendScore":14,"compositeScore":16,"rankGlobal":9,"rankLanguage":9,"license":17,"archived":18,"fork":18,"defaultBranch":19,"hasWiki":20,"hasPages":18,"topics":21,"createdAt":9,"pushedAt":9,"updatedAt":38,"readmeContent":39,"aiSummary":40,"trendingCount":14,"starSnapshotCount":14,"syncStatus":41,"lastSyncTime":42,"discoverSource":43},93266,"vox-director","Alisa0808\u002Fvox-director","Alisa0808","Turn one topic into a finished Vox-style paper-collage explainer\u002Fad video — automated end to end on Atlas Cloud + ffmpeg. An agent skill.",null,"Python",158,19,1,0,25,58.9,"MIT License",false,"main",true,[22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37],"ai","ai-video","claude-code","claude-skill","collage-video","explainer-video","ffmpeg","generative-ai","llm","motion-graphics","python","text-to-video","tts","video","video-generation","vox","2026-07-22 04:02:08","\u003Cp align=\"right\">\u003Cb>English\u003C\u002Fb> · \u003Ca href=\"README.zh.md\">简体中文\u003C\u002Fa>\u003C\u002Fp>\n\n# 🎬 Vox Director\n\n**Turn one topic into a finished Vox-style paper-collage explainer \u002F ad video — script, collage keyframes, motion, voice-over, music and captions, all automated.**\n\nAn **agent skill** that runs end to end on the [Atlas Cloud](https:\u002F\u002Fwww.atlascloud.ai\u002F?utm_source=github&utm_campaign=vox_director) API + local `ffmpeg`, usable by any coding agent (Claude Code, Codex, etc.). You give it a one-line topic; it gives you an `mp4`.\n\n![License: MIT](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-black.svg) ![Powered by Atlas Cloud](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fpowered%20by-Atlas%20Cloud-ff5a1f.svg) ![Agent Skill](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FAgent-Skill-d97757.svg)\n\n\u003Cdiv align=\"center\">\n\nhttps:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Fed08d230-7bcb-4b48-a17d-23c079208f9f\n\n\u003Cb>▶ \"The evolution of Chinese civilization\" · 30s\u003C\u002Fb>\n\n\u003C\u002Fdiv>\n\n\u003Ctable>\n  \u003Ctr>\n    \u003Ctd width=\"33%\">\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F216cd62f-6314-456c-94cf-1090b8559a22\">\u003Cimg src=\"assets\u002Fthumbs\u002Ffootball.jpg\" width=\"100%\" alt=\"How football conquered the world\">\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd width=\"33%\">\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F561788b1-5615-4828-b3f8-b24ae5ad7bcd\">\u003Cimg src=\"assets\u002Fthumbs\u002Fmexican.jpg\" width=\"100%\" alt=\"Mexican street food\">\u003C\u002Fa>\u003C\u002Ftd>\n    \u003Ctd width=\"33%\">\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002Ff69f072f-f50a-41ba-9e66-7ed0aae4ddc0\">\u003Cimg src=\"assets\u002Fthumbs\u002Fmoney.jpg\" width=\"100%\" alt=\"A brief history of money\">\u003C\u002Fa>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n    \u003Ctd align=\"center\">\u003Csub>Football history · 60s\u003C\u002Fsub>\u003C\u002Ftd>\n    \u003Ctd align=\"center\">\u003Csub>Mexican street food · 60s\u003C\u002Fsub>\u003C\u002Ftd>\n    \u003Ctd align=\"center\">\u003Csub>A brief history of money · 60s\u003C\u002Fsub>\u003C\u002Ftd>\n  \u003C\u002Ftr>\n\u003C\u002Ftable>\n\n\u003Cp align=\"center\">\u003Csub>\u003Cem>▶ more films — click any thumbnail to play\u003C\u002Fem>\u003C\u002Fsub>\u003C\u002Fp>\n\n---\n\n## What it is\n\nThe look is the modern editorial **paper-collage** popularized by Vox explainers: hand-cut paper cut-outs, torn edges, tape, halftone dots, newspaper clippings, bold flat color per beat, big cut-out headlines — brought to life with motion, a narrator, music and captions.\n\n## How it works\n\nOne topic flows through one script per stage, all driven by a single `beats.json` per project:\n\n```\ntopic\n  │\n  ├─ 1. beat map        pick a narrative arc → write beats.json      ◀── GATE 1: you approve the beat map\n  ├─ 2. style bake-off  render the same beat in 3–4 themes           ◀── GATE 2: you pick the look by eye\n  ├─ 3. keyframes       one collage poster per beat  (nano-banana-2)\n  ├─ 4. motion          animate each poster          (gemini-omni-flash i2v)\n  ├─ 5. voice + music   one narrator (xai\u002Ftts) + BGM (minimax\u002Fmusic)\n  ├─ 6. assemble        ffmpeg: concat, duck music under VO, burn captions + watermark\n  └─ final.mp4\n```\n\nTwo ideas make or break the result, and the skill is built around both:\n\n1. **The look is born in the image step.** Each beat is a finished collage *poster*. All the collage DNA (torn paper, cut-outs, halftone, headline text) lives in that image — if the poster isn't a rich collage, nothing downstream saves it.\n2. **The motion is added after.** By default an AI video model animates the whole poster (the \"living poster\" path). For dramatic *piece-by-piece* assembly, an optional local keyframe engine cuts the poster into parts and drives them frame-by-frame (no content filters, pixel-exact — great for real people).\n\nTwo human decision gates keep you in control (approve the beat map; pick the style); everything else is automated.\n\n## Models (verified on Atlas Cloud)\n\n| Job | Model |\n|---|---|\n| Keyframe \u002F collage poster | `google\u002Fnano-banana-2\u002Ftext-to-image` |\n| Animate (non-real content) | `google\u002Fgemini-omni-flash\u002Fimage-to-video` |\n| Animate (**real people \u002F brands**) | `kwaivgi\u002Fkling-video-o3-pro\u002Fimage-to-video` |\n| Narration | `xai\u002Ftts-v1` |\n| Music | `minimax\u002Fmusic-2.6` |\n| Cut out an element (advanced path) | `youchuan\u002Fv8.1\u002Fremove-background` |\n\nModel IDs drift — the skill fetches the live list from `GET https:\u002F\u002Fapi.atlascloud.ai\u002Fapi\u002Fv1\u002Fmodels` before running.\n\n## Install\n\nThis is an **agent skill** — it works with any coding agent that can read a workflow and run scripts (Claude Code, Codex, …). Claude Code auto-discovers it as a skill; other agents read [`AGENTS.md`](AGENTS.md) → [`SKILL.md`](SKILL.md).\n\n**Option A — from this repo:**\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002FAlisa0808\u002Fvox-director.git ~\u002F.claude\u002Fskills\u002Fvox-director\n```\n\n**Option B — from the packaged skill:** download [`vox-director.skill`](vox-director.skill) and install it via your Claude skills UI.\n\nThen set your Atlas Cloud API key (get one at [atlascloud.ai\u002Fconsole\u002Fapi-keys](https:\u002F\u002Fwww.atlascloud.ai\u002Fconsole\u002Fapi-keys?utm_source=github&utm_campaign=vox_director)):\n```bash\nexport ATLASCLOUD_API_KEY=\"sk-...\"\n```\n\n## Quick start\n\nJust ask your coding agent, with the skill installed:\n\n> *\"Make me a Vox-style collage video introducing Mexican street food — English, 16:9, 15 seconds.\"*\n\nThe agent will draft a beat map for your approval, run a style bake-off for you to pick from, then generate keyframes → motion → voice → music and assemble `out\u002F\u003Cproject>\u002Ffinal.mp4`.\n\n## Requirements\n\n- A **coding agent** — Claude Code, Codex, or similar\n- **Atlas Cloud** API key\n- **ffmpeg** + **ffprobe** (`brew install ffmpeg`)\n- **Python 3** with **Pillow** (`pip install pillow`) — for caption\u002Fwatermark overlays\n\n## What's in the box\n\n```\nSKILL.md              the skill (English) — the workflow the agent follows\nSKILL.zh.md           the same skill in Chinese\nAGENTS.md             entry point for non-Claude agents (Codex, …)\nreferences\u002F           the creative engine\n  prompt-guide.md       the LOOK layer — prompt structures, vocab & 8 theme presets\n  beat-layer.md         14 narrative arcs + hook\u002Fpacing + shot patterns\n  voices.md             xai\u002Ftts voice roster — pick a voice_id per language\u002Ftone\n  models-and-gotchas.md every API \u002F ffmpeg gotcha, already solved\n  local-engine.md       the advanced element-level motion engine\nscripts\u002F              one script per pipeline stage\nexamples\u002F             ready-to-run beats.json examples\nassets\u002F               the showcase film\n```\n\n## Credits\n\nInspired by the collage-ad workflows of **[Stav Zilber](https:\u002F\u002Fx.com\u002FStavZilber)**, **[rom1trs](https:\u002F\u002Fx.com\u002From1trs)** and **[Higgsfield](https:\u002F\u002Fx.com\u002Fhiggsfield_ai)**, and by **[Vox](https:\u002F\u002Fwww.vox.com)**'s explainer visual language.\n\nBuilt end to end on **[Atlas Cloud](https:\u002F\u002Fwww.atlascloud.ai\u002F?utm_source=github&utm_campaign=vox_director)** — one prompt, one film.\n\n## License\n\n[MIT](LICENSE) © 2026 Atlas Cloud\n","Vox Director 是一个自动化生成 Vox 风格纸艺拼贴解说视频的 AI 代理技能。它接收单行主题输入，端到端完成脚本生成、分镜海报设计（纸艺拼贴风格）、关键帧动画、TTS 语音合成、背景音乐添加及字幕烧录，最终输出 MP4 视频。项目基于 Atlas Cloud API 与本地 FFmpeg 协同工作，融合 LLM（如 Claude）、多模态模型（如 Gemini i2v）和生成式 AI 技术，强调视觉风格一致性与叙事节奏控制。适用于知识科普、品牌宣传、教育内容快速原型制作等需高效产出高辨识度视觉化解释视频的场景。",2,"2026-07-15 02:30:03","CREATED_QUERY"]